You ask a question whose answer is not in your facts
The answer says she does not know and names what is missing. That is an exact answer, not a failure: reading “unknown” beats assuming.
Asking a question about your own business and getting an answer drawn from your data, with the date of every piece of information used.
An answer to a question about your business, written from the facts you approved (your services, your hours, your rules) and your instructions — not from general knowledge. “Which hours did I give for Saturday”, “which rule did I set for emergencies”: those are owner questions, and they have exact answers.
Every claim comes with what it rests on and the date it was recorded. If the information does not exist, it shows as unknown rather than being filled in with a guess — the single most important difference between this screen and a chat bot.
When the answer calls for an action — call someone back, send a confirmation, offer another slot — you get a draft. Nothing goes out until you confirm it.
The screen also reads well on a phone: that is why it exists, rather than forcing you to read a dashboard between jobs.
“Which hours did I give for Saturday”, “which rule did I set for emergencies”, “what is the last instruction I published”. No need to phrase it like a search: it is a question, not a filter.
Each element of the answer shows where it came from: which approved fact or instruction, and the date the information was recorded. The source and the date appear next to the element.
What she does not know is displayed as such. That is often the most useful signal: missing information points to an instruction to write or a field nobody collects.
If an action is proposed, it arrives as a draft with its recipient and its text. You confirm, you edit, or you discard. The right to contact that person is checked before sending is possible.
The facts you approved and your published instructions. Nothing else: she does not reach for general knowledge about your trade to plug a gap, and she does not read another business's data.
Every element of her answer points back to the fact that carries it, with its date. You find that fact in your workspace and correct it if needed. That is what makes the answer usable for deciding something, rather than merely pleasant to read.
When the answer calls for an action, it arrives as a draft: recipient, channel, wording. You read it, correct it, confirm it. Until you confirm, nothing leaves — the screen does not take it upon itself to write to one of your customers.
Before sending, the right to contact the person is checked. A withdrawn consent, or a reply window closed on the channel's side, blocks the send and shows which of the two applies. The draft stays available for another channel, or for a phone call you make yourself.
The answer says she does not know and names what is missing. That is an exact answer, not a failure: reading “unknown” beats assuming.
It shows as unknown. Connecting the software makes it readable, and the connections screen states what the connection genuinely permits. Without a connection, add the fact to your instructions if the answer has to be available.
Each person sees what their role and scope allow. Someone attached to a single site gets an answer about that site, with the others absent, including from any total.
Each send stays a separate draft, with its own check on the right to write. Nothing goes out in one gesture: it is slower, and it is what avoids a message reaching someone who asked to stop receiving them.
The screen opens in your phone's browser, with the same limits: what she does not know stays marked unknown.
The information may live somewhere other than your workspace — in a notebook, or in software that is not connected. Add it to your instructions or connect the software concerned.
An old fact stays active, and its date is shown. Correct it in your workspace: she relies on the version you approve.
Either the person asked not to be contacted again, or the channel window is closed. The draft stays, the send is blocked: the screen says which of the two applies.
The questions about this screen.
No, and the difference is visible to the naked eye: every statement names the data behind it and its date, and what is missing shows as unknown instead of being filled in.
No. This screen is for your team, according to the roles you granted. What the receptionist says to your customers comes from your published instructions, not from your internal data.
The ones you would answer by digging through your instructions: your hours, your services, your transfer rules, what you asked her to say or not to say.
An owner's question, phrased the way you would say it out loud, gets a better answer than a keyword.
Every element carries the date the fact was recorded. An old fact stays active, and the date is there to remind you.
She prepares the message; you send it. The right to contact the person is checked before sending is possible, and the screen says what blocks it if anything does.
Yes, and that is why the screen exists: to find an hour, a rule or an instruction between two jobs, without reading a full dashboard.
Every task has its own.
The two screens that go with it.
If this article does not settle your case, write to the address below: a person reads it and answers you in writing. Give your business name, what you were trying to do and, for a call, its date and time.
[email protected]Contact us: what to write depending on your request